1. Field of the Invention
The present invention relates to an optical cracked-grain selector that optically determines and selects for removal cracked grains in material rice grains such as brown rice, polished rice and the like.
2. Description of Related Art
Conventionally, an optical cracked-grain selector apparatus that optically detects and selects grains of rice having one or more cracks that penetrate to the interior of the grain (hereinafter “cracked grain” or “cracked grains”) is known, such as those disclosed in JP2005-265519A and JP3642172B. In general, the above-mentioned crack in the rice grain usually extends in a direction substantially perpendicular to a longitudinal direction of the rice grain. As shown for example in FIG. 10, a conventional optical cracked-grain selector 100 comprises a slanted chute 200 for pouring the material rice grains downward, and optical detecting means 300 and selection means 400 disposed near a bottom end of the slanted chute 200 at positions along a downward trajectory G of fall of the material rice grains. The optical detecting means 300 has an emitter 300a disposed on one side of the trajectory G of fall that emits a line-like laser beam of light toward an optical detection position P on the trajectory G of fall and a CCD (charge-coupled device) camera 300b disposed on the other side of the trajectory G of fall that detects light at the optical detection position P. Such an optical cracked-grain selector 100 sends the material rice grains down the slanted chute 200, irradiates the material rice grains with the optical detecting means 300 when they pass the optical detection point P on the trajectory G of fall so as to capture the light passed through the grains with the CCD camera 300b, processes signals using cracked grain identification means 500 separately provide and identifies cracked grains based on received light data, and selects the identified cracked grains for removal using the selection means 400.
However, the optical cracked-grain selector 100 described above has the following problem. Specifically, the cracked grain identification means 500 identifies a total image (total visual image) of each grain of rice based on the received light data and identifies a cracked grain whenever it detects linear dark shades of data corresponding to cracks in the image of each grain thus identified. However, each grain of rice has an embryonic portion and sometimes also surface cracks in the skin (hereinafter “scratches”), and it is known that the embryonic portions and the scratches adversely affect identification accuracy when identifying cracked grains. In other words, when the embryonic portion and scratches are present in the rice grain, they show up as dark shadows just like cracks, and for that reason normal grains having no cracks are wrongly identified as cracked grains, causing a decline in product yield.